See also: IRC log
<trackbot> Date: 24 October 2006
<scribe> Scribe: pauld
minutes from the 17th approved
<Yves> http://www.w3.org/2006/10/17-databinding-minutes.html
WG welcomes Hugo Haas from Yahoo! as an observer
discussion of interop event
yves: having people in a room would help socialisation and crafting of tests
jonc: getting end users involved would be a challenge
pauld: getting vendors to participate has been difficult
... organising around the W3C Workshop on Enterprise Services might help
hugo: an interop event sounds appealing, not too high cost from vendors and may help build test cases
yves: also a mini-TP around then in Boston, could be co-located
pauld: OK, let's consider an event early 2007 in Boston, we can always cancell if we get no interest
pauld: built some "traffic lights" to schema validate our instance documents, and if the detected patterns match the expected status:
http://www.w3.org/2002/ws/databinding/examples/6/09/
http://www.w3.org/2002/ws/databinding/patterns/6/09/
pauld: can we change the normative XPaths without the change being considered substantive?
yves: we will at least have the text?
pauld: we have very little text as the XPaths are normative
yves: sounds like we would have to go back to Last Call
pauld: will ensure the XPaths are correct, then!
pauld: will call for a vote to go to Last Call for Basic on Tuesday
pauld: can we publish a first WD of Advanced?
vlad: sections 2.1 and 2.2 are missing
pauld: OK, that looks editorial, I'll fix./
pauld: hearing no objections, we'll plan to publish first WD of Advanced alongside LC of Basic
pauld: worked on the test client, still Perl, still works on separate files
hugo: took a brief look, seems sensible approach
otu: took a look at WCF, svcgen no longer generates a server databinding, only interfaces. Suggested we use Visual Studio 2005, but you want a scripted interface.
otu: also started to look at Python ZSI, struggling with lack of documentation
pauld: WCF still generates a client databinding no?
otu: we can turn this inside out
yves: may get time to work on XMLUnit later this week
pauld: will publish test suite page and ping WSO2 for assistance
pauld: there's a stylesheet on the bottom of the patterns page, should work on Schemas and WSDLs (though not tried many WSDLs)
pauld: we still need to follow schema import and include and detect UTF-8 etc
yves: if we think the pattern detector we have works, but will still look at an alternative approach
pauld: could still be useful, but let's make it lower priority than XMLUnit
pauld: anticipate having to write some non-XSLT code to check some of our non-XPath based assertions
pauld: submitted evidence on why it's not Basic pattern, but is it Advanced?
vladislav: SAP implementations now support this
pauld: so it sounds advanced, let's see if Priscilla comes up with any new info on why it might be a bad idea before closing it.
vladislav: we have had
limitations mapping Unicode characters to language
constructs
... also we have length limitations
hugo: do you have a sense of a "safe" character set?
<Yves> http://www.w3.org/2002/ws/databinding/issues/10/
pauld: difficult - Fortran or COBOL might
offer a lowest common denominator, but no obvious lowest subset exists, but we seem to be moving towards Python's subset
... plan to close this issue based on testing next
week
<scribe> ACTION: vladislav to report on naming convention supported by his toolkit [recorded in http://www.w3.org/2006/10/24-databinding-minutes.html#action01]
<trackbot> Created ACTION-87 - Report on naming convention supported by his toolkit [on Vladislav Bezrukov - due 2006-10-31].